skills/zhinkgit/embeddedskills/keil/Gen Agent Trust Hub

keil

Pass

Audited by Gen Agent Trust Hub on Sep 15, 2026

Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes the Keil uVision executable (UV4.exe) using the subprocess module to perform build, clean, and flash operations.
  • Evidence: In scripts/keil_build.py, the run_uv4 function constructs a command list [uv4_exe, flag, str(project_path), "-j0", "-o", str(log_file)] and executes it using subprocess.run.
  • Security Note: The implementation uses a list for the command arguments rather than a single string, which is a best practice that prevents shell injection vulnerabilities.
  • [INDIRECT_PROMPT_INJECTION]: The skill reads and parses XML-based Keil project files (.uvprojx, .uvproj).
  • Evidence: scripts/keil_project.py and scripts/keil_build.py use xml.etree.ElementTree to extract TargetName, OutputDirectory, and OutputName.
  • Security Note: While these files are external data, the skill only extracts specific structural metadata required for the build process, posing a low risk of influencing the agent's behavior maliciously.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 15, 2026, 11:14 AM